Stop–start practice on your own — gradual steps, no scoreboard

For two or three sessions, note the time from starting stimulation to the start of ejaculation. Keep the window consistent (stimulation begins → ejaculation begins, pauses not counted) and take the range, not an average. Two or three readings are rough data — that's expected, and it's enough. The number is a reference point for your own trend, not a verdict.

If clock-watching winds you up, don't time it: mark each session high / medium / low by feel, or count breaths. Anxious monitoring isn't a neutral measuring tool — in one lab experiment, adding a distraction task suppressed physical arousal, and self-critical self-monitoring during sex is linked to difficulties. A noisier baseline you can live with beats a precise one that wrecks the practice.

If you want a number to steer by, an example format: aim a bit longer than last time — a minute more is plenty. That's an illustration, not a prescription; the size of the step is yours to set, and 'a bit longer' by feel counts too.

Expect a non-linear path. Sleep, stress, alcohol, and medications all move your response, so runs of shorter sessions sit inside normal variance. 'Direction' means the trend across weeks — a week that looks like week one is information, not failure, and there is no finish line where the practice is 'done'.

When you feel the pull toward finishing, stop all stimulation — hands off, nothing moving — for a few seconds until the wave subsides, then resume. This is the stop–start technique from behavioral therapy tradition: stop near the threshold, resume once excitement ebbs, repeat.

The stop isn't a glitch in the practice — the back-and-forth is the practice. Each cycle is one repetition of noticing where your edge is and letting it pass.

If stopping alone doesn't cool things fast enough, the classic companion is the squeeze technique from the Masters & Johnson tradition: alongside stopping, apply firm, steady pressure with thumb and fingers to the penis for a few seconds.

Reviews list both techniques but standardize neither the spot, nor the pressure, nor the duration — so calibrate to your own body: steady pressure, brief, and pain means it's wrong, not 'working'. Skip the folk mechanism about squeezing blood out of the penis; nothing supports it. And know the frame: these techniques are described in clinical, often partner-assisted contexts — practicing solo is your extension of the tradition, with one small single-group study of a related technique behind the solo form.

The urge to rush is partly an attention problem: focus collapses onto the finish line. An anchor gives it somewhere else to be. Options in equal standing — an imagined scene you can stretch out, the sensation of your own breathing, or the feel of your body generally. An imagined partner scene is one tool, not a requirement; if fantasy isn't available or pleasant for you, breath and body-sense are full citizens.

Keep the tone light: an anchor is something you try for a few seconds, not a skill to master.

Arousal dips mid-practice are normal. Waiting out a softer stretch is part of the skill — 'waiting beats caving into the urge to just finish' is a reasonable working rule.

Ending the session is equally legitimate: put it away, another session exists. A dropped erection is data about today, not a disqualification.

Masturbation and partnered sex run under different conditions: difficulties show up as milder during masturbation across surveys, and men with premature ejaculation specifically report longer latency while masturbating. Whether solo practice carries over hasn't been tested directly — in either direction — so hold both possibilities.

When you bring the technique into partnered situations, recalibrate from scratch. Researchers read the context gap as higher arousal and performance pressure — their interpretation, since the surveys measured symptoms, not those factors. Either way the conditions differ: bring the noticing skill, drop the expectations, and learn this context's version of your edge as if new. If performance pressure itself is the loudest thing in the room, that's its own article — linked below.