SYNTH-74 · Guide
Swim training for triathletes and fitness swimmers
Anchor every pace to a tested CSS, swim often and short rather than rarely and long, put technique before the main set, and do not expect the bike or run to keep the swim fit (one n=4 model); the swims per week, session design, send-offs and gear rules are practice, not evidence
The question it answers
How to test CSS and what it anchors; swims per week and session structure; main sets and send-offs; technique versus fitness for a weak or a competent swimmer; pull buoy, paddles, fins; open water and wetsuit; the swim taper; what swim fitness gives the bike and run. Whole-race preparation is in SYNTH-52 (sprint/Olympic) and SYNTH-57 (70.3/Ironman).
What the base actually holds
CSS is a real threshold concept, validated once. In nine trained male college swimmers in a flume, critical velocity from swims to fatigue at six speeds (mean 1.166 m/s) correlated with the speed at the onset of blood lactate accumulation (r = 0.949) and with 400 m freestyle speed (r = 0.864) (REF-874, B, test validation).The two-distance pool version: in well-trained competitive swimmers CSS from 200- and 400-m tests sat 3.2 % above the speed of a 30-min swim, not statistically different (REF-931, B; no sample size given).CSS is not the lactate steady state: in 8 well-trained swimmers it was 92.7 % of maximal aerobic speed against 88.3 % at MLSS — correlated (r = 0.87), not interchangeable (REF-932, B).
Efficiency separates swimmers from triathletes. Propelling efficiency was 61% in elite swimmers versus 44% in triathletes, power output changes dramatically with training, and distance per stroke is a fairly good indicator of efficiency (REF-534, B+, 1992 narrative review). The primary study: 6 competitive swimmers vs 5 triathletes at equal power input, distance per stroke 1.23 m vs 0.92 m (REF-933, B; a group contrast, not an intervention).
Intensity domains must be defined. A narrative review stresses accurately defined intensity domains for designing swim training and introduces an "optimal swimming technique speed" tied to how stroke rate and stroke length change in exhaustive swims (REF-535, B+; no numbers in its abstract; that stroke length falls under fatigue is from the full text, not checked).
Swim fitness is swim-only. In four elite triathletes modelled over 40 weeks, no cross-transfer was identified between swimming and cycling or running (REF-785, B, case series; the swim model fitted poorly, r = 0.37).
More metres did not beat harder metres in daily-training swimmers. Two 1.5 h sessions a day for 6 wk gave collegiate men no more than one (REF-934, B, matched groups); in 18 elite swimmers season improvement correlated with mean training intensity (r = 0.69), not volume or frequency (REF-935, B, observational). Neither tests swims per week.
Default (what to prescribe first)
For a healthy adult-onset triathlete who can swim 400 m front crawl without stopping:
1
Test CSS, then pace everything from it. Not from a study; the product's convention: 400 m and 200 m all-out in the same session after a full warm-up, with easy swimming between them, CSS = 200 m ÷ (400 m time − 200 m time). The 200 + 400 m version is validated in well-trained swimmers only (REF-931, B); in adult triathletes it is transfer.
CSS anchors targets (% of CSS) and swim load. Heart rate is not used to prescribe swimming (product rule; no entry validates swim heart-rate zones).
2
Three swims a week; two is the floor. Not from a study; the base's own practice: 3 × 30-60 min for a time goal, 2 to hold the swim; a fourth short swim before a longer one. Inference: frequency is preferred to length because the gap to swimmers is in efficiency (REF-534) and a skill is practised by repetition; no entry compares swims per week in triathletes.
3
One session shape. Not from a study; the base's own practice: warm-up 300-600 m easy with a few short pickups → technique block 200-400 m (drills in 25-50 m pieces, each followed by plain swimming) → one main set → 100-200 m easy. The 1,000-1,500 m in REF-290 (B+, narrative review) is a pre-race warm-up for competitive pool swimmers, not a session rule; warm-up logic → SYNTH-66.
4
One CSS set a week, one aerobic set, one mixed speed/technique set. Not from a study; the base's own practice (menu below).
Main-set menu and send-off logic
All of this is house practice; no entry compares swim sets in adult triathletes.
CSS set (repeats with rest, not steady swimming): repeats of 100-400 m at 98-102% of CSS speed with 10-20 s rest, 800-2,000 m in total depending on race distance (for example 8-12 × 100 m, 5-6 × 200 m, 4-5 × 400 m). CSS is the upper edge of threshold work: above the lactate steady state and above 30-min speed (REF-932, REF-931).
Speed: 25-50 m fast with long rest, 200-400 m in total. In 18 national-level 13-year-old swimmers a 2 × 10 × 25 m set at 100 m pace gave higher velocity with lower lactate, heart rate and RPE than 5 × 50 m, but lower distance per stroke (REF-653, B, acute crossover, no training outcome; transfer to adults is inference).
Send-off = target swim time + planned rest, rounded to 5 s. If rest shrinks below ~5 s or pace drifts more than ~2-3 s per 100 m, end the set or ease the send-off; do not lower CSS from one bad set.
Decision
| Situation | Do this | Source and how sure |
|---|---|---|
| Cannot yet swim 400 m continuously | No CSS test. Short repeats of 25-50 m with generous rest, most of the session technique; lessons or a coach on deck. Test when 400 m is continuous | House practice; no entry studies adult learners |
| Weak adult-onset swimmer (far slower in the water than on land for the same fitness) | 3 short swims; about a third to a half of each session technique; CSS set short (for example 6-8 × 100 m); distance per stroke at a fixed pace as the marker | REF-534 B+, REF-933 B (61% vs 44% efficiency; distance per stroke as the skill marker; the author's advice is technique first); the shares are practice |
| Competent swimmer (ex-club, or steady pace and stroke count through a CSS set) | Technique shrinks to 200-300 m of the warm-up; volume and CSS/aerobic sets carry the progress | REF-534 B+ (power output rises dramatically with training); the split is practice |
| Sprint / Olympic triathlete | 2-3 swims; CSS repeats of 100-200 m; some fast 25-50 m; race-start practice (draft-legal: front-pack speed → SYNTH-52); otherwise swim below time-trial effort | SYNTH-52 (practice). In 9 highly trained male triathletes in laboratory sprint triathlons, a swim at 80-85% of time-trial velocity gave a faster bike and overall time than one at 98-102% (REF-940, B, crossover; no drafting, no magnitudes); Olympic distance: narrative only (REF-834, C+) |
| 70.3 / Ironman triathlete | 2-3 swims, one aerobic set with longer repeats (300-800 m) at about 88-95% of CSS, the menu's aerobic band; race the swim controlled | Sets are practice. In nine male Ironman triathletes the swim had the highest heart rate of the day (149.2 vs 137.1 and 136.2 bpm), and a larger swim-to-bike heart-rate gap went with slower later stages (REF-836, B, observational, one race) |
| Fitness swimmer, no race | Same test and session shape; 2-4 swims; sets by preference; other strokes as easy metres | House practice; no entry studies fitness swimmers |
| Pool swimmer with an event | Shape the hard fraction to the event | Highly trained sprinters followed polarized/threshold, middle-distance threshold/pyramidal, long-distance pyramidal distributions (REF-650, B+, systematic review of 9 mostly observational studies); transfer to amateurs is inference |
| Hours are short | Cut swim length, not swim frequency; do not drop the swim for the bike | REF-785 B (n=4 elites: nothing crossed into swimming); frequency-over-length is practice |
Pull buoy, paddles, fins, kick
One entry measures fins and one paddles in a single sprint; none measures pull buoys or a training effect of any gear. The rules are house practice.
Pull buoy: for pull sets and for drills where sinking legs hide the fault. Cap at roughly a quarter of weekly metres for a weak swimmer: it masks the body-position problem that needs fixing.
Paddles: only with a sound catch and pain-free shoulders; small paddles, short repeats. Shoulder pain → remove paddles first.In eight female swimmers sprinting 25 m of front crawl, small (116 cm(2)) and large (268 cm(2)) paddles raised propelling efficiency, stroke length and swimming velocity and lowered stroke rate significantly, which the authors say might argue against paddle training, particularly with large paddles (REF-976, B; acute mechanics, no training outcome, no shoulder data).Small rather than large is that entry's inference.
Fins: for drills and kick sets so the drill can be done at a sensible body position; not for the CSS set. At comparable aerobic speeds kicking with fins cost about 40 % less energy (REF-937, B; kicking only) — the same pace is an easier set.
Kick: 100-300 m a session for body position; no large kick sets for triathletes.
Inference: gear changes pace, so a pull or fins set is not evidence that CSS moved, and load computed from pace versus CSS overstates such sets.
How to progress / periodise
Retest CSS at the start and end of a block, roughly every 8-12 weeks, or when the send-offs say so. Not from a study; the base's own practice (the same interval, also labelled practice, as SYNTH-53): same pool length, same warm-up, same order, fresh. A CSS set held with rest to spare at a lower stroke count is the sign to retest.
Progress one thing at a time (practice): main-set metres, then less rest, then pace.
Strength: across 27 studies in competitive swimmers, swim plus strength training seemed better than swimming only, with similar gains of about 2-2.5% whatever the method (REF-651, B+, systematic review without pooling; transfer to adult triathletes is inference). Gym programme → SYNTH-44 / SYNTH-43.
Open water and wetsuit: wetsuit use and swim drafting improve performance and efficiency relative to pool swimming (REF-538, B+, 2002 narrative review). Over 400 m a wet suit cut 8 triathletes' time by 19 s and left 8 competitive swimmers' unchanged (REF-938, B; pool — do not scale to race distance).Passive drag fell 21% and 20% at 0 and 50 cm behind a leader's toes, and 6% and 7% beside, 50 and 100 cm back from the hands (REF-939, B, 11 swimmers, towing; the beside figures from six of them).Not from a study; the base's own practice: from ~8 weeks out swim open water in the race wetsuit every 1-2 weeks (sighting, starts, swimming on feet); pool sets continue; set race effort from rehearsals, not from pool CSS.
Taper → [SYNTH-34]. Across 27 studies the most efficient taper was 2 weeks, volume cut 41-60%, intensity and frequency kept (REF-107, A, mixed sports). In 8 highly trained pool swimmers a 4-week taper improved competition performance by 2.32% (REF-536, B, uncontrolled, taper content not described). Not from a study; the base's own practice: keep the number of swims, shorten the main sets, keep a few CSS-pace repeats, and do not stop swimming in race week.
When NOT to use it / red flags
Open water is a safety problem, not a comfort one. Water-temperature limits rest on little evidence, acclimatised swimmers cannot judge their own thermal state, and cardiac events appear to be the primary cause of death in open-water events (REF-654, B+, narrative review). Repeated cold exposures reduce discomfort (REF-562, B+, review).Never swim open water alone; chest pain, palpitations or breathlessness out of proportion → out of the water and a doctor (practice line of REF-654).
Shoulder pain that persists into the next session: stop paddles and hard pull, cut volume, see a physiotherapist (practice; the base holds no swimmer's-shoulder entry).
Do not prescribe swim sets by heart rate, and do not copy bike or run zones into the pool (product rule; no entry validates swim HR zones).
Do not present CSS as "the pace you can hold for X minutes" — REF-874 shows ranking agreement with lactate markers, nothing about duration.
Do not promise a stated gain from drills — no entry tests a drill.
What the evidence does and does not show
Shows: what is listed under "What the base actually holds", plus a relatively harder swim going with slower later legs in nine Ironman men (REF-836), a wetsuit gain in triathletes but not in swimmers (REF-938), less drag on a leader's feet (REF-939) and a faster sprint triathlon after a sub-maximal swim (REF-940).
Does not show: that the two-distance test holds in adult triathletes, or that CSS is a steady-state pace; that technique work beats fitness work in adult-onset swimmers (no intervention); how many swims a week are needed; any training effect of drills, pull buoy, paddles or fins; any send-off rule; open-water pace from pool CSS; anything specific to women, masters or fitness swimmers.
Evidence gaps
Pull buoy; any training effect of paddles or fins; paddles and shoulder load — REF-976 is one maximal 25 m in 8 women.
Swimmer's shoulder, swim heart rate versus land, adult learners, masters swimmers — no entries.
How this entry was checked and changed
Why this level
(CSS rests on three B validation studies in trained swimmers; technique on one B group contrast and B+ narrative reviews; wetsuit, drafting, fins, paddles (acute, n=8) and race-swim intensity on single B studies; transfer on an n=4 elite model, B; no entry tests swims per week in triathletes, drills, paddle training, pull buoys or send-offs — all practice)
Content correction 2026-09-27
Revision 2026-10-01: folded REF-931 and REF-932 (Dekerle 2002, 2005). The two-distance CSS test is no longer 'validated by no entry': it is validated in well-trained swimmers (3.2 % above 30-min speed) and stays a transfer for adult triathletes.Because CSS sat above the maximal lactate steady state (92.7 % vs 88.3 % of maximal aerobic speed, n = 8), the main-set heading 'CSS / threshold' is now 'CSS set (repeats with rest, not steady swimming)' with CSS described as the upper edge of threshold work; the 98-102% band, rests and totals remain practice. The aerobic set carries the same point as an inference.
Revision 2026-10-01: folded REF-933 (Toussaint 1990, the primary study behind 61% vs 44%), REF-934 and REF-935 (volume vs intensity in daily-training swimmers — stated as not a test of swims per week), REF-937 (fins: kicking only), REF-938 (wetsuit: 19 s over 400 m in 8 triathletes, none in swimmers — replaces 'no verified number for the wetsuit gain'), REF-939 (drafting distance) and REF-940 (Peeling 2005, sprint-triathlon swim intensity, now the source of the sprint/Olympic pacing row ahead of the narrative review).
Revision 2026-10-01: six Evidence-gap lines narrowed to what is still missing (adult triathletes, swims per week, paddles and pull buoy, race-distance wetsuit gain, open-water drafting, longer races). Level unchanged at B-: every new entry is a small B study in trained swimmers or triathletes, and the week, the sets, the send-offs and the gear rules are still practice.
Verification 2026-10-01: all 25 cited entries opened in the KB.md snapshot; levels, designs, n and numbers next to each refId match the entry head blocks (1.166 m/s, r 0.949/0.864, 3.2 %, 92.7/88.3 %, r 0.87, 61/44 %, 1.23/0.92 m, r 0.37, r 0.69, 80-85 vs 98-102 %, 149.2/137.1/136.2 bpm, ~40 %, 2-2.5 %, 19 s, 21/20/6/7 %, 41-60 %, 2.32 %).
Verification 2026-10-01: no unlabelled prescription found: the Default steps, the main-set menu, send-offs, gear rules, retest interval, open-water and taper practice each carry a practice / product-rule / inference label in the line or the section header. Full-text-unchecked content of REF-535 is labelled as such.
Verification 2026-10-01: the 70.3 pace band (90-97% of CSS) vs the menu bands, and the stale CSS/wetsuit lines in SYNTH-52, are owner decisions; no pair written. The lead says 'two narrative reviews' while the body cites more (REF-534, 535, 538, 290, 654, 834).
Verification 2026-10-01: no SOURCE flags: every cited entry has a DOI or PMID; REF-931 and REF-934 give no sample size and the SYNTH says so.
2026-10-01: the 70.3/Ironman row's 90-97% of CSS matched neither band of this card's own menu (CSS set 98-102%, aerobic 88-95%); it now uses the menu's aerobic band. All three are practice.
Revision 2026-10-01: REF-976 (Gourgoulis 2008, B, eight female swimmers, single maximal 25 m) folded into the paddles rule: paddles raised propelling efficiency, stroke length and velocity and lowered stroke rate — the authors' own doubt about paddle training, particularly large paddles.The house rule 'small paddles' now carries the entry's inference label; 'sound catch, pain-free shoulders, short repeats' stay practice (no shoulder data in the entry). No default changed; level C+ unchanged.
Revision 2026-10-01: lead, Level parenthetical and Evidence gaps updated (paddles: an acute study exists, a training effect does not); the 'candidate studies named from memory' line removed because no candidate remains in the list.