Purpose: To study the change of the tumor control probability (TCP) caused by variation in the fractional dose from the interplay between tumor motion and intensity modulated radiation therapy (IMRT). Methods: A numerical simulation was performed to simulate the dose perturbation induced by the interplay. 107 tumor clonogens were divided into 1, 10 and 100 sub‐volumes that received an expected fractional dose of 2 Gy with dose variations following either a normal distribution or a sine distribution. The variation was generated independently to each sub‐volume. Two different groups of alpha/beta values were used to represent fast and slow proliferation tumors, respectively. TCP was calculated for 10,000 independent histories using the linear quadratic model, with or without the repopulation of the tumor cells. Results: The interplay did not significantly change the TCP for tumors with 1 sub‐volume. On the other hand, a significant decrease in the TCP was observed in a more probable scenario with different dose variation in different tumor subvolumes. Higher radiation doses are required to achieve the same TCP for 95% of the patients. Approximately, for alpha/beta=1 1.9 Gy, an additional 0.69 Gy and 0.37 Gy are required for an additional 0.1 Gy of uncertainty with the normal and sine distributions, respectively; for alpha/beta=3 Gy, the additional doses are 0.73 Gy and 0.41 Gy. Tumor repopulation exacerbates the problem for the normal distribution by increasing additional doses required to 0.92 Gy (alpha/beta=11.9 Gy) and 1.05 Gy (alpha/beta=3 Gy) for every additional 0.1 Gy of uncertainty in the delivered dose. Conclusions: Interplay between tumor motion and IMRT contributes negligibly to the total physical dose but significantly to the tumor control probability. A higher dose is needed to achieve the same TCP if random variation in the fractional dose is present.