A full night stabilizes; a nap can let performance slip
The clearest contrast comes from comparing a nap with a night's sleep. In healthy older adults and in people who had suffered a stroke, motor performance decreased after a nap but stabilized after a night's sleep [3]. That is a direct observation of the same people under two sleep conditions, which makes the difference hard to attribute to anything but the sleep itself.
This matters because "sleep" is often treated as a single ingredient. A 90-minute nap and a full night differ in how much of each sleep stage a person gets, and the stabilization seen overnight may depend on that fuller architecture. The nap result is a reminder that a short sleep is not simply a smaller dose of the same medicine.
Who benefits most, and who may not
Age and brain health change the picture. Older adults are described as showing impaired sleep-dependent motor memory consolidation, and one proposed explanation is dysfunction in sleep spindles, the brief bursts of brain activity characteristic of sleep [3]. When researchers tried to boost consolidation in older adults by replaying a sound tied to a learned sequence during a nap, performance on the reactivated sequence was no better than on the control sequence [1]. Notably, both the associated and control sounds changed brain responses, which leaves open the possibility that auditory stimulation helped in a general, non-specific way rather than through targeted reactivation [1].
Hand dominance also shapes the outcome. In a study of 48 young adults learning a serial color-matching task, all groups improved during training, but at the 24-hour retest the groups that had practiced with their dominant hand performed best, while those who practiced with the non-dominant hand were weaker [2]. The design bundles handedness and practice hand together, so it supports the idea that dominant-hand practice plus sleep gives the strongest consolidation, without isolating which factor does the work.
What inside sleep seems to separate strengthening from flexibility
The most detailed clue comes from the timing of spindles. Spindles do not fire evenly; they cluster into "trains," and they can lock onto slow oscillations, the slow waves that sweep the sleeping brain. In one study, immobilizing an arm after motor learning changed the timing of this coupling for spindles grouped in trains but not for isolated spindles [4]. The authors report that spindle trains may promote skill-specific strengthening of motor memories, while isolated spindles may instead create conditions of memory instability that favor generalization [4]. Those are proposed roles, not settled functions, but they offer a concrete reason why the same night of sleep could sharpen one skill and loosen another.
Spindle density also appears protective in a specific setting: in older adults and stroke survivors, higher spindle density was associated with less memory loss after a nap [3]. After stroke, longer spindle duration was linked to better consolidation and interpreted as a possible compensatory mechanism to stabilize memory [3]. These are associations within particular groups, so they describe who tends to hold onto motor memory rather than proving that spindles cause the retention.
When sleep adds nothing at all
Sometimes the learning itself already does the work. When a new motor sequence is highly compatible with one learned earlier, a "schema," people integrate it quickly. Two experiments tested whether sleep is needed for that benefit: one compared a 90-minute nap with wakefulness, the other a 12-hour interval containing nocturnal sleep with one containing only wakefulness. In both, motor sequence performance (reaction time and accuracy) did not differ between sleep and wake groups, and in the nap experiment sleep features such as non-REM duration, spindles, and slow waves did not correlate with post-sleep performance [5].
The practical reading is that sleep's contribution is not automatic. When the new skill slots neatly into an existing motor framework, the framework may carry the consolidation, leaving little for sleep to add. When the skill is novel, or when the sleeper is older or recovering from stroke, sleep's role becomes more visible, and so does the difference between a nap and a night.
Sources used in this answer
Targeted memory reactivation during post‐learning sleep does not enhance motor memory consolidation in older adults
Replaying a sound associated with a learned motor sequence during a nap did not improve performance on that sequence in older adults compared with a control sequence, though both sounds modulated brain responses.
Effect of Hemispheric Asymmetry and Sleep on Motor Memory Consolidation
In young adults, all groups improved with training, but at a 24-hour retest those who practiced with their dominant hand showed the best motor memory performance, while non-dominant-hand practice groups performed worse.
Enhancing sleep-dependent motor memory consolidation with non-invasive brain stimulation in healthy older adults and individuals after stroke
Motor performance decreased after a nap but stabilized after a night's sleep in healthy older adults and stroke survivors, and higher sleep spindle density was associated with less nap-related memory loss.
Temporal sleep spindle clustering and slow-oscillation coupling in motor memory consolidation and generalization
Spindle trains may promote skill-specific strengthening of motor memories, while isolated spindles may create memory-instability conditions that favor generalization.
Sleep does not influence schema-facilitated motor memory consolidation
Sleep after learning a motor schema did not improve motor sequence performance compared with wakefulness in either a nap or a nocturnal-sleep experiment.