Estimating blood pressure (BP) from pulse arrival time (PAT) by image-based (skin video) photoplethysmography (iPPG) is of increasing interest due to the non-contact method advantage (over cuff-based methods) and potential for BP measurement to be built into portable devices such as mobile phones. The relationship between pulse transit time extracted from iPPG has been investigated during stable BP. The sensitivity of beat-to-beat iPPG-PAT to dynamic changes in BP has not been explored. This study investigated the correlation between iPPG-PAT and diastolic BP (DBP) during 1-minute seated rest and 3-minute isometric handgrip exercise. 15 healthy participants (9 female, 34±13 years) were recruited. Video was recorded from subjects' faces at 30 frames per second using a standard web-camera with simultaneous measurement of the electrocardiogram and noninvasive finger BP. The iPPG waveform was from the averaged green channel intensity of regions of the forehead or cheek. PAT was calculated from the R-wave ofthe electrocardiogram to the foot of the iPPG or finger BP waveform respectively for direct comparison. Handgrip exercise caused a steady increase in DBP (75±9 to 87±13 mmHg, p<0.001). Beat-to-beat iPPG-PAT and DBP was negatively correlated (mena ±SE -1.33±1.70 ms/mmHg, P=0.0024) as was finger-PAT (mean ±SE -0.5S ±0.39 ms/mmHg, P<0.001). The proportion of individual significant negative regression slopes between DBP and finger-PAT and between DBP and iPPG-PAT was not significantly different. Despite high variability of the correlation between iPPG-PAT and DBP among subjects, iPPG-PAT can track dynamic changes in BP.