<h1>Manuscript</h1> This repository includes stimuli and analyses codes for the following manuscript: **Joensen, B. H., Bush, D., Vivekananda, U., Horner, A. J., Bisby, J. A., Diehl, B., Miserocchi, A., McEvoy, A. W., Walker, M. C., & Burgess, N. (2022). Hippocampal theta activity during encoding promotes subsequent associative memory in humans.** <h1>Abstract</h1> Hippocampal theta oscillations have been implicated in associative memory in humans. However, findings from electrophysiological studies using scalp electroencephalography (EEG) or magnetoencephalography (MEG), and those using intracranial EEG (iEEG) are mixed. Here we asked ten pre-surgical epilepsy patients undergoing iEEG recording, along with 21 participants undergoing MEG recordings, to perform an associative memory task, and asked whether hippocampal theta activity during encoding was predictive of subsequent associative memory performance. Across the iEEG and MEG study, we observed that theta power in the hippocampus increased during encoding, and that this increase differed as a function of subsequent memory, with greater theta activity for pairs that were successfully retrieved in their entirety compared to those that were not remembered. This result helps to clarify the role of theta oscillations in associative memory formation in humans, and further, demonstrate that findings in epilepsy patients undergoing iEEG recordings can be extended to healthy participants undergoing MEG recordings <h1>Repository</h1> This repository contains the file "Stimuli.xslx" and the folders *iEEG* and *MEG* - The file *Stimuli.xsls* contains the stimuli used in both the iEEG and MEG study. - The folder *iEEG* contains code required to replicate the iEEG results. - The folder *MEG* contains code required to replicate the MEG results. <h3>iEEG</h3> The code needed for computing estimates and performing analysis of power are *EstPow.m*, *CmpStat.m*, and *TstRsp.m*, respectively. - Estimates of time-frequency power are computed using *EstPow.m*. Estimates are computed using the SPM12 function ‘spm_eeg_morlet’ (available at https://www.fil.ion.ucl.ac.uk/spm/software/spm12/). This code outputs individual .m files. - *CmpStat.m* performs a cluster-based permutation analysis (implemented in Fieldtrip, available at https://www.fieldtriptoolbox.org/) on time-frequency power. - *TstRsp.m* performs *F*- and *t*-tests on estimates time-frequency power (computed using *EstPow.m*) across subsequent memory performance. Time and frequency bands of interest are derived from time-frequency clusters identified using *CmpStat.m*. - The code *CmpTilt.m* is used for computing and comparing ‘spectral tilt’ across subsequent memory. The codes *CmpRgn.m* and *CorPow.m* are used to estimate and compare/correlate theta power across contacts in the hippocampus, amygdala, and lateral temporal lobe. <h3>MEG</h3> The codes needed for (*i*) estimating source-reconstructed power (*ii*) smoothing source-reconstructed power, (*iii*) and generating the second level model (*F*-contrast) for comparison of subsequent memory performance are *Sme_FrstLvl_SpcEst.m*, *Sme_FrstLvl_Smth.m*, and *Sme_ScndLvl_SpcEst.m*, respectively. These are all implemented in SPM12 (available at https://www.fil.ion.ucl.ac.uk/spm/software/spm12/) and use the DAiSS toolbox (available at https://www.fil.ion.ucl.ac.uk/spm/ext/#DAiSS). The code *Sme_ExtractBetas.m* can be used to extract source-reconstructed estimates of power for a given mask. This code requires the MarsBar toolbox (available at https://www.fil.ion.ucl.ac.uk/spm/ext/#MarsBar). <h1>Comments</h1> Please direct any comments to Bárður H. Joensen, bardur.hofgaard.joensen@psyk.uu.se. We cannot provide support for you to adapt the code to your own data.