For infants with acute progressive hydrocephalus, invasive drainage of cerebrospinal fluid (CSF) is performed until a ventriculo-peritoneal shunt can be inserted. Surrogate markers of intracranial pressure (ICP) may help optimise the timing of invasive procedures. To assess whether RI with/without fontanel compression helps distinguish between infants with normal (<5 cmH2O), mild (5–11 cmH2O), and moderate (>11 cmH2O) ICP elevation, 74 ICP measures before/after CSF removal and 148 related Doppler measures of the middle cerebral artery were assessed. Higher RI was associated with fontanel compression, elevated ICP, and their interaction (all p < 0.001). Without compression, differences in RI were observed between normal and moderate (p < 0.001) and between mild and moderate ICP elevation (p = 0.033). With compression, differences in RI were observed for all pairwise comparisons among normal, mild, and moderate ICP elevation (all p < 0.001). Without compression, areas under the receiver-operating characteristic curve for prediction of mild and moderate ICP elevation were 0.664 (95% confidence interval (CI), 0.538–0.791; p = 0.020) and 0.727 (95% CI, 0.582–0.872; p = 0.004), respectively, which improved to 0.806 (95% CI, 0.703–0.910; p < 0.001) and 0.814 (95% CI, 0.707–0.921; p < 0.001), respectively, with compression. RI with fontanel compression provides improved discrimination of infants with absent, mild, and moderate ICP elevation.