An innovative method of reducing external phosphorus (P) loading to lakes uses engineered systems to treat lake inflows with aluminum sulfate (alum). In this study we used a series of jar tests to examine the optimal alum dose and mixing regime to remove P from Matthiesen Creek, an important external source of P to Jameson Lake. Matthiesen Creek is a good candidate for alum treatment because the creek runs year round, and the majority of P in the spring-feed creek is in the form of bioavailable dissolved P that can be efficiently captured in alum floc. The mixing regimes in this study mimicked a range of possible treatment scenarios that relied on natural turbulence in the creek or conventional mechanical mixing, and presumed the discharge of alum floc either directly to the lake or to an on-shore settling basin. Jar tests showed that an alum dose of 5 mg-Al/L was sufficient to decrease P from around 0.13 mg-P/L to below 0.02 mg-P/L for most mixing regimes. For all mixing regimes, doses of up to 20 mg-Al/L did not depress pH below the recommended minimum pH of 6. Flash mixing prior to low-intensity mixing did not enhance P removal over low-intensity mixing alone, but flash mixing alone resulted in lower levels of P removal from creek water. Jar testing with a mixture of alum-treated creek water and lake water showed that lake waters tended to inhibit P uptake by alum floc. This, combined with the fact that high pH favors the formation of the aluminate ion which could exhibit chronic toxicity to aquatic biota, suggests that discharge of alum solids directly to the lake should be avoided. We recommend an engineered inflow treatment system on Matthiesen Creek that maintains an alum dose of 5–10 mg-Al/L under moderate mixing conditions (Gt of 1,000–3,000) with alum floc collected in an on-shore settling basin.