Manipulation of nanoparticles with atomic force microscopes (AFMs) has been under development for a decade and is now well established as a technique for prototyping nanodevices and for other applications. Until now, the manipulation process for particles with sizes of a few nanometers has been very labor intensive. This severely limits the complexity of the structures that can be built. Particle sizes on the order of 10 nm are comparable to the spatial uncertainties associated with AFM operation, and a user in the loop has been needed to compensate for these uncertainties. This paper addresses thermal drift, which is the major cause of errors for AFMs operating in ambient conditions. It is shown that drift can be estimated efficiently by using Kalman filtering techniques. This approach has firm theoretical foundations and is validated by the experimental results presented in this paper. Manipulation of groups of 15-nm particles is demonstrated under program control, without human intervention over a long period of time, in ambient air and at room temperature. Coupled with existing methods for high-level motion planning, the manipulation capabilities introduced here will permit assembling, from the bottom up, nanostructures that are more complex than those being built today with AFMs. Note to Practitioners-Nanomanipulation with scanning probe microscopes (SPMs) has potential applications in nanodevice and system prototyping, or in small-batch production if multitip arrays are used instead of single tips. However, SPM nanomanipulation is still being used primarily in research labs. A major obstacle to its wider use is the labor and time involved in the process. These are largely due to spatial uncertainty in the position of the tip (which is analogous to a robot's end effector) relative to the sample being manipulated. Today, a skilled user is needed to determine where the tip is and to correct manipulation errors due to inaccurate positional estimates. The major cause of this spatial uncertainty is thermal drift between the tip and the sample. At the time scales relevant to manipulation, the drift can reach values comparable to the size of the objects, especially if these are below ∼10 nm. The techniques discussed in this paper compensate for the dri- ft and enable automated manipulation, with associated savings in time and labor, and increased complexity of the resulting structures.