OBJECTIVE The determination of gait improvement after lumbar puncture (LP) in idiopathic normal-pressure hydrocephalus (iNPH) is crucial, but the best time for such an assessment is unclear. The authors determined the time course of improvement in walking after LP for single-task and dual-task walking in iNPH. METHODS In patients with iNPH, sequential recordings of gait velocity were obtained prior to LP (time point [TP]0), 1-8 hours after LP (TP1), 24 hours after LP (TP2), 48 hours after LP (TP3), and 72 hours after LP (TP4). Gait analysis was performed using a pressure-sensitive carpet (GAITRite) under 4 conditions: walking at preferred velocity (STPS), walking at maximal velocity (STMS), walking while performing serial 7 subtractions (dual-task walking with serial 7 [DTS7]), and walking while performing verbal fluency tasks (dual-task walking with verbal fluency [DTVF]). RESULTS Twenty-four patients with a mean age of 76.1 ± 7.8 years were included in this study. Objective responder status moderately coincided with the self-estimation of the patients with subjective high false-positive results (83%). The extent of improvement was greater for single-task walking than for dual-task walking (p < 0.05). Significant increases in walking speed were found at TP2 for STPS (p = 0.042) and DTVF (p = 0.046) and at TP3 for STPS (p = 0.035), DTS7 (p = 0.042), and DTVF (p = 0.044). Enlargement of the ventricles (Evans Index) positively correlated with early improvement. Gait improvement at TP3 correlated with the shunt response in 18 patients. CONCLUSIONS Quantitative gait assessment in iNPH is important due to the poor self-evaluation of the patients. The maximal increase in gait velocity can be observed 24-48 hours after the LP. This time point is also best to predict the response to shunting. For dual-task paradigms, maximal improvement appears to occur later (48 to 72 hours). Assessment of gait should be performed at Day 2 or 3 after LP.