Cuckoo hashing is a highly practical dynamic dictionary: it provides amortized constant insertion time, worst case constant deletion time and lookup time, and good memory utilization. However, with a noticeable probability during the insertion of n elements some insertion requires Ω(log n) time. Whereas such an amortized guarantee may be suitable for some applications, in other applications (such as high-performance routing) this is highly undesirable. Recently, Kirsch and Mitzenmacher (Allerton '07) proposed a de-amortization of cuckoo hashing using various queueing techniques that preserve its attractive properties. Kirsch and Mitzenmacher demonstrated a significant improvement to the worst case performance of cuckoo hashing via experimental results, but they left open the problem of constructing a scheme with provable properties. In this work we follow Kirsch and Mitzenmacher and present a de-amortization of cuckoo hashing that provably guarantees constant worst case operations. Specifically, for any sequence of polynomially many operations, with overwhelming probability over the randomness of the initialization phase, each operation is performed in constant time. Our theoretical analysis and experimental results indicate that the scheme is highly efficient, and provides a practical alternative to the only other known dynamic dictionary with such worst case guarantees, due to Dietzfelbinger and Meyer auf der Heide (ICALP '90).