The gradient system is an essential component in MR imaging. Unfortunately, there remain many technical factors that inevitably cause distortions in the realized gradient magnetic field. Therefore, it is challenging in practice to realize the actual gradient field as prescribed, which results in image artifacts (e.g., blurriness, ringing, or phase error). In this study we propose a new gradient measurement technique based on single point imaging (SPI), which allows simple, rapid, and robust measurement of k-space trajectory. In the proposed technique, 1D SPI data is acquired by linearly scaling the prescribed gradient waveform along each axis. The field of view (FOV) in SPI at phase encoding time delay (t_p) is $$FOV(t_p)=πN_p/(γ\int_{0}^{t_{p}}{}G(τ)dτ)$$, where N_p is the number of phase encoding, γ is the gyromagnetic ratio, and G(τ) is a maximum amplitude of phase encoding gradient at time delay τ. Therefore, the FOV changes with phase encoding time delay, exhibiting a zoom-in (time decreasing FOV) or zoom-out (time increasing FOV) effect as shown in Figure 1. Figure 2-a shows an example of SPI encoding to measure a trapezoidal readout gradient. After 1D SPI data is acquired, relative FOV scaling factors are found, which directly reflect the relative k-space trajectory with respect to the reference encoding time, t_r, as follows: $$FOVscale(t) = FOV(t)/FOV(t_r) = k(t)/k(t_r)$$, where t denotes a phase encoding time delay, and k(t) is a k-space position in the unit of cycle m^-1 at encoding time, t. There are two possible approaches to estimate the FOV scaling factor: an image domain or an k-space domain approach. In the image domain approach, relative scaling of the object size between two time delays is used to directly estimate an FOV scaling factor. In the k-space domain approach, relative linewidth of k-space can be measured, which is the inverse of the FOV scaling factor. This can be formulated as a nonlinear optimization problem for which we used negative linear correlation and L₂ norm as cost functions for the image or k-space domain approach, respectively. Note that the 1D profile in the image domain has more resolution when the FOV is small (Figure 2-b), while the k-space profile shows a broader line-shape when the FOV is large (Figure 2-c), and vice versa. Therefore, the FOV scaling factor estimated by two different approaches are combined using a merging filter as shown in Figure 2-d. To evaluate the propose method, three experiments were conducted: (1) 3D radial UTE human brain imaging (IRB approved) on a 3T MR scanner (GE MR750) with TE=90µs, TR=3.3ms, and # of radial spokes=80,000. Gradient measurement was performed on the x, y, and z axes with N_p=401 and scan time=4sec. (2) 2D spiral phantom imaging on a 1.5T MR scanner (GE HDxt) with spiral arms=16, readout points/arm=990, FOV=24cm, spatial resolution=2x2mm, TE=13ms, TE=2.4ms. Gradient measurement with Np=801 was performed in two different ways for comparison: (a) all 16 different pairs of x and y gradients measured (extensive measurement=167sec) and (b) 2 pairs of x and y gradients were measured and reproduced to estimated trajectories for all 16 arms using linear combination (quick measurement=42sec). (3) 3D multi-echo Cartesian GRE phantom imaging for IDEAL¹ with and without ramp sampling on a 3T MR scanner (GE Signa PET/MR). Parameters were spatial resolution=2x2x2mm, TEs/TR with ramp sampling=0.80,1.5,2.1ms/3.65ms and TEs/TR without ramp sampling=0.99,1.9,2.8ms/4.78ms. Gradient measurement was performed with Np=401 in the readout direction and a scan time of 1.5sec.Figure 3-a shows the measured UTE trajectory in the physical x, y, z axis, and prescribed trajectory, and a zoomed-in view. Figure 3-b and c show UTE images reconstructed with the prescribed trajectory and the measured trajectory. The image reconstructed with the measured trajectory shows good quality with no visible imaging artifact such as ringing. Figure 4-a shows the spiral image reconstructed using fully measured trajectory. Figure 4-b shows the spiral image reconstructed with quick gradient measurement. Figure 5 shows the multi-echo IDEAL images reconstructed to a conventional (5-a) and ramp-sampled with measurement (5-b) trajectory. The ramp-sampled trajectory allowed a 25% reduction in scan time and incorporating the measured trajectory into reconstruction improved water-fat separation without the use of any additional phase correction.The proposed SPI-based gradient measurement technique allows robust and fast (particularly when TR is short) gradient measurement with no special hardware, and can be performed with very minor modifications to the targeted pulse sequence to provide substantial improvements in image quality. This technique will also likely be useful for linear system modeling of gradient waveforms².