Small non-axisymmetric magnetic fields are known to cause serious loss of stability in tokamaks, leading to loss of confinement and abrupt termination of plasma current (disruptions). The best known examples are the locked mode and the resistive wall mode. Understanding of the underlying field anomalies (departures in the hardware-related fields from ideal toroidal and poloidal fields on a single axis) and the interaction of the plasma with them is crucial to tokamak development. Results of both locked mode experiments (Scoville J.T. and La Haye R.J. 2003 Nucl. Fusion43 250) and resistive wall mode experiments (Garofalo A.M., La Haye R.J. and Scoville J.T. 2002 Nucl. Fusion42 1335) done in DIII-D tokamak plasmas have been interpreted to indicate the presence of a significant anomalous field. New measurements of the magnetic field anomalies of the hardware systems have been made in DIII-D. The measured field anomalies due to the plasma shaping coils in DIII-D are smaller than previously reported (La Haye R.J. and Scoville J.T. 1991 Rev. Sci. Instrum.61 2146). Additional evaluations of systematic errors have been made. New measurements of the anomalous fields of the Ohmic heating and toroidal coils have been added. Such detailed in situ measurements of the fields of a tokamak are unique. The anomalous fields from all the coils are one-third the values indicated from the stability experiments (Garofalo et al 2002, Scoville and La Haye 2003). These results indicate limitations in the understanding of the interaction of the plasma with the external field. They indicate that it may not be possible to deduce the anomalous fields in a tokamak from plasma experiments and that we may not have the basis needed to project the error field requirements of future tokamaks.