TR and TE in MRI

Two timing parameters build every basic MRI contrast. TR decides how much tissues recover between pulses; TE decides how much they decay before the echo is read. Here is how, with typical values and the scan-time arithmetic.

By the ScanFlo Editorial Team · Updated 3 October 2026

TR (repetition time) is the time between successive excitation pulses for the same slice, and TE (echo time) is the time from the excitation pulse to the peak of the echo that is read out. TR sets how much T1 recovery tissues get between pulses; TE sets how much T2 decay happens before the signal is measured. Short TR with short TE gives T1 weighting, long TR with long TE gives T2 weighting, and long TR with short TE gives proton density.

TR: who has recovered?

After an RF pulse tips magnetisation over, each tissue's longitudinal magnetisation grows back at its own T1 rate. TR is how long you wait before tipping it again (Radiopaedia, Repetition time). Cut that wait short and only fast-recovering tissue, such as fat, has much to give, while slow tissue, such as CSF, stays dark. That difference is T1 contrast. Wait long enough and everything recovers, and the T1 differences disappear.

How short is short? Elster's rule of thumb is that "long" means about three to five times the tissue's T1 or T2, and "short" means much less than it (Questions and Answers in MRI). With white matter T1 around 790 ms at 1.5 T, a TR near 500 ms is short; a TR of 3,000 ms or more is long.

TE: who has decayed?

Once the spins are tipped, the transverse signal fades at each tissue's T2 rate. TE is how long you wait before reading it (Radiopaedia, Echo time). Read early and every tissue still has most of its signal, so T2 makes little difference. Read late and short-T2 tissue has faded while long-T2 tissue such as CSF and oedema still shines. That is T2 contrast. The trade-off is signal: a longer TE always means less of it, so T2-weighted images are noisier for the same time.

The four combinations

Sources: the four-way scheme is from Elster; T1 TR example from the University of Michigan fMRI course; TE ranges from Radiopaedia; PD TR from Radiopaedia, PD weighted spin echo images; T2 spin echo values from the IAEA handbook. These are spin echo values at 1.5 T. Gradient echo uses much shorter TR and TE, with the flip angle taking over part of TR's job; see gradient echo.

Worked example: what TR does to the clock

For a conventional 2D spin echo, scan time is TR × phase-encoding lines × averages (Sung, UCLA M219 lecture). With 256 phase-encoding lines and one average:

That is why conventional T2 spin echo has been replaced by fast spin echo, which collects 16, 32 or more lines in each TR (IAEA handbook, §14.5.5). It is also why a long TR is less costly than it looks: other slices are excited during each slice's dead time, so a longer TR fits more slices into the same scan (IAEA, §14.5.6). The cost moves rather than vanishes. More echoes per TR blur the image slightly, and more slices per TR mean more RF heating.

Common mistakes

• Lengthening TR to buy signal on a T1 scan. More TR means more recovery for every tissue, and the T1 contrast you wanted fades along with the noise.
• Pushing TE up on a T2 scan without checking SNR. Past the point where fluid is clearly bright, a longer TE mostly adds noise.
• Reusing 1.5 T timings at 3 T. T1 values are longer at 3 T, so the same TR gives a different amount of T1 weighting.
• Treating TR and TE as independent of everything else. Echo train length, number of slices and SAR limits all push back on the TR you can set.
• Forgetting that fast spin echo TE is an effective TE: each line of k-space has its own echo time, and contrast follows the lines at the centre.

The 3 T effect is from Stanisz et al. (2005); the different TE per k-space line in fast spin echo is described in the IAEA handbook, §15.4.3. For how the two parameters show up on real images, read T1 vs T2 weighting and spin echo.