A major benefit of hydropower is its ability to respond quickly to fluctuating electrical loads. However, the sharp changes in discharge caused by this practice have detrimental environmental effects downstream. This study investigated the effects of hydrograph shape on attenuation of regulated pulsed flow events by first categorizing, then modeling the downstream movement of representative pulses on the upper Tuolumne River below Holm Powerhouse in the Sierra Nevada mountains of California. This system was managed by a public utility and produced flow pulses primarily for hydroelectricity generation and/or whitewater recreation. Operations were highly influenced by a system-wide "Water First" policy, which prioritized drinking water supply and quality over other beneficial uses. Pulses were therefore associated with a spectrum of time scales, from predetermined schedules decided far in advance to hydropeaking operations responding to real-time demands. We extracted underlying hydrograph shape patterns using principal component analysis on individual pulsed flow events released from 1988-2012 (n=4439). From principal component loadings, six shape categories were determined: rectangular, front-step, back-step, goalpost, centered tower, and other. The rectangular and stepped shapes were the most frequent, composing 62% and 24% of total events, respectively. The rectangular shape was often produced by 'standard' hydropeaking or recreational releases, while the stepped shapes were often used for water conservation or were recreational flows bordered by periods of electricity generation. The stepped shape increased in occurrence after the "Water First" policy took effect in 1993 and dominated two drier years (2007 and 2009). After categorization by shape, magnitude and durational indices were used to fabricate representative pulsed flow events. Attenuation of these representative pulses was then modeled using a 1D hydraulic model of 42 river km prepared in HEC-RAS. As no operational measures or physical structures existed within the system to counter the adverse effects of pulsed flow events, natural attenuation was the only potential major mitigation agent. However, model results demonstrated a clear durational threshold for representative pulses (~ 3-5 hrs) over which the degree of attenuation of ramping rates and peak discharge approached a limit. These thresholds were unique to the study reach and were dependent upon river morphology, bed characteristics, and flow rates. Increasing baseflows did not necessarily increase attenuation of pulses, most likely due to minimal increases in bed friction forces in this fairly steep and confined channel. Simulations of front and back-step representative pulses showed trade-offs between attenuation of peak magnitudes and steepness of ramping rates. Finally, a range of rising ramping rates were shown to steepen downstream above initial rates due to the study reach's channel morphology. Reshaping pulses to be more ecologically benign at all points downstream was infeasible if the system was required to maintain current electricity production and recreational service levels.