Weekly Training Signal — April 10, 2026
Every week, the Weekly Training Signal identifies one signal from training data and gives you the practical protocol that follows from it. This week: training frequency per muscle group.
The signal
One of the most debated questions in gym culture remains training frequency per muscle group. The classic bro-split, one muscle group per day across 5 days, was long the dominant format in gyms. Its logic: maximize volume on a single muscle group in one intense session.
Meta-analyses on muscle protein synthesis tell a different story.
The science
Muscle protein synthesis (MPS), the biological mechanism behind muscle growth, peaks in the 24 to 48 hours following training, then returns to baseline. Training a muscle group only once per week leaves 5 to 6 days without a protein synthesis stimulus.
Available meta-analyses converge on this: at equivalent total weekly volume, training each muscle group twice per week produces significantly more hypertrophy than training it once per week.
For the step from 2x to 3x per week, the data is less definitive. Most studies show similar results between 2x and 3x for most people, with a slight positive trend for 3x in advanced athletes. But the marginal benefit decreases as frequency rises, while recovery demands increase.
The recommended protocol
For the large majority of people: 2x per week per muscle group is the balance point between stimulus and recovery.
Program structures that achieve this frequency:
Push/Pull/Legs (PPL) 6 days: each muscle group is hit twice per week, rotating Push / Pull / Legs / Push / Pull / Legs.
Upper/Lower 4 days: upper and lower body sessions alternating, each pattern twice per week.
Full body 3 days: each muscle group stimulated 3x per week with lower per-session volume. Effective option for people with limited time.
If you're on a bro-split 1x/week and getting good results: keep going. But if you've plateaued, increasing frequency to 2x with the same total volume is probably the most direct lever to restart progress.